50M : Jitter

This is mine as of 30 seconds ago, I type very slowly

[img]Download Failed (1)[/img]

[qasdfdsaq]

Originally Posted by kwikbreaks

Admittedly they've either improved my local network or the neighbourhood torrent freak now has the entire output of 20C Fox et. al. on Bluray and has retired to watch them all but I used to see garbage results from TBB when my local connection was OK and now they are fine. It seemed more likely to me that the Linx loading had reduced somehow although whether that was down to increased capacity or re-routing something else I've no idea.

Can't have been LINX loading or rerouting because only end-user connections saw any change.

Core nodes, CMTS, and everything else saw no change. If it were core network or interchange improvements anything on VM's network would be responding faster and not just end users on cable modems.

[kwikbreaks]

Surely the ping packets traverse the entire VM network? What about my improved speedtests from TBB?

[Chrysalis]

Also that not everyone is seeing improvements.

Whenever I did ping analysis of my VM connection the following pattern was consistent.

No or almost no jitter right up until my CMTS.
Occasionally the CTMS itself would have some jitter but usually the jitter started at my ip.

This is harder to diagnose on outbound trace's because then the jitter shows on every hop but easy on inbound. So whatever VM's peering issues are they werent the cause of VM's jitter.

Also there has been VM graphs that have been like the improved ones posted that have existed for many months, my neighbour is on a upstream channel that seems to be very lightly utilised or maybe that channel already had the new configuration which seems to have been rolled out now to some other users.

[kwikbreaks]

I'm in a poor position to judge.

When I first noticed problems with TBB I was on 50 and there was no noticeable local congestion. Then came the local problems which I put down to a single 100/10 sale as the onset was so fast plus you'd sometimes see a huge improvement for a short time then really quickly back to dross (for 50) so I downgraded to 10 which I pretty much always got.

I first noticed better TBB charts maybe a month or so back - somewhere around the time I ordered Infinity. Then Openreach naffed me around so I decided to give VM another chance and upgraded to 60/3. Over this period lots of people have been reporting improvements + R36 which I'm now on offers less jitter.

Given that lot I've really no idea why I now see headline rate speedtests from TBB despite it all still going through Linx as it always has.

[craigj2k12]

as ive just posted in another thread discussing the same subject I think VM have made p2p throttling far more aggressive, these could be the improvements we are seeing